-
-
Notifications
You must be signed in to change notification settings - Fork 28.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Using Voice/Assist causes the websocket to get overloaded leading to HAOS stability issues #92307
Comments
just wanted to add more information on this, when I use Home Assistant cloud with these settings (running b8): asking the voice assistant same command for the kitchen lights as above, it works amazingly well and fast (unbelievably fast, wow!), but It still throws the logs, but not as much as when using local voice.
Using the local voice and it will log in the thousands as mentioned above. |
Same problem with my HA (2023.5.1) using assist on RPI4, before Whisper the STT was ok .. now this feature is unusable ! |
Same experience on Home Assistant 2023.5.1, Voice Assist works properly (on Nabu Casa Cloud from phone) but logged some 260 |
Hi, Just an update on this. It still does it when I use Frontend (i.e. the microphone in the browser), however, when I use VOIP: no issues and no error messages in the core logs as above. So, this is probably something related to frontend. |
I am getting a lot of these messages every time I try using Voice Assistant via the Android App.
|
For people experiencing this, what browser and platform are you using? And are you browsing the website via Nabu Casa Cloud or local or proxy? |
This is happening to me when I use Voice Assistant in my HA Android App on my Samsung phone with Android 13. |
Since my posts are kind of scattered I can provide some more information in a summary here. It does not happen when I use VoIP (via grandstream and a phone) and it does not happen when using the Atom5 Echo (via ESPHome). And this is true for both local and cloud-based setups, no issues and no errors. The problem occurs when using frontend ways such as (Windows OS and Chrome browser) and app (Android). In these two cases, the error messages approach the thousands when I use local-based processing and the hundreds when I use home assistant cloud. Hope that provides more information. |
Can anything be done about this? I have the same errors flowing in my v2023.7.0. |
I just started playing with this on v2023.7.1 and I'm getting these errors in my log as well....when using the Android mobile app. Also connecting via Nabu Casa. |
Can you share the logs? |
I have much less of these errors in
|
For me, running this project (https://github.com/synesthesiam/homeassistant-satellite) triggers a constant stream of these in the logs:
HASS v2023.10.2 |
+1 |
Hey there @home-assistant/core, @synesthesiam, mind taking a look at this issue as it has been labeled with an integration ( Code owner commandsCode owners of
(message by CodeOwnersMention) conversation documentation |
Same issue here with satellite `Logger: homeassistant.components.assist_pipeline.pipeline Unexpected error during intent recognition error on satellite side: |
Here too, when stopping the 'homeassistant-satellite' service on all satellites, the error subsides. |
+1 |
Came here to see what's up with the logs after installing wake word. Logger: homeassistant.components.websocket_api.http.connection [140034822700480] Received binary message for non-existing handler 3 Home Assistant container running in docker with separate containers piper, whisper and porcupine (rhasspy/wyoming-piper, rhasspy/wyoming-whisper, rhasspy/wyoming-porcupine1) |
Keeping this alive in 2024!
|
I have the same problem: `Logger: homeassistant.components.websocket_api.http.connection [140195976403264] Received binary message for non-existing handler 9` Frontend + microphone. ESPHome + micronoffal does not work for me at all. :( Core |
There hasn't been any activity on this issue recently. Due to the high number of incoming GitHub notifications, we have to clean some of the old issues, as many of them have already been resolved with the latest updates. |
The problem is still there. HA Core: Errors
|
The problem
Hi,
Running HAOS (2023.5.0b4) on RPi4 4Gb on 500GB SSD (usb boot) and using HA in Chrome (Windows 10 computer) to use Assist.
Working with @bdraco, what we did:
The profiler service was started, and then the Assist dialogue box was opened and I voiced my command by pressing the microphone key "Turn on the Kitchen Light", it responded and then replied back did the action, then I said "Turn off the kitchen light" and likewise it responded and did the action. Both commands were successful. Shown here:
However, the logs would indicate otherwise. During this time, errors were logged in the thousands (over a few minutes) during the above dialogue:
My voice assistant settings are:
And the callgrind files here:
callgrind.out.1682882817681.txt
profile.1682882817681262.cprof.txt
Reveal some issues:
Where "20% of the event loop time is logging 1000x of the messages"
During this time, memory usage increased and CPU usage spiked (see the end of the traces):
In other separate callgrinds, similar issues are also seen:
This may also be related to this issue.
What version of Home Assistant Core has the issue?
2023.5.0b4
What was the last working version of Home Assistant Core?
not sure
What type of installation are you running?
Home Assistant OS
Integration causing the issue
Assist
Link to integration documentation on our website
No response
Diagnostics information
No response
Example YAML snippet
No response
Anything in the logs that might be useful for us?
No response
Additional information
No response
The text was updated successfully, but these errors were encountered: